Hum Genet (1991) 87 :377-388

63 Springer-Verlag 1991

Review article

The phenylketonuria locus: current knowledge about alleles and mutations of the phenylalanine hydroxylase gene in various populations David S. Konecki and Uta Lichter-Konecki Universitats-Kinderklinik, Im Neuenheimer Feld 150, W-6900 Heidelberg, Federal Republic of Germany Received December 17, 1990 1 Revised February 10, 1991

Summary. The hyperphenylalaninemic disorders of classic phenylketonuria (PKU), mild phenylketonuria, and hyperphenylalaninemia (HPA), result from a deficiency of the hepatic enzyme phenylalanine hydroxylase (PAH) or its cofactor (tetrahydrobiopterin). Use of the complementary DNA of this enzyme has allowed the establishment of a restriction fragment length polymorphism (RFLP) haplotype-analysis system. This haplotype analysis system provides the means for determination of mutant PAH alleles in most affected families and is the basis for mutational analysis of the PKU locus. This review is focused on two major areas of current PKU research: (1) the use of DNA haplotype analysis in the study of the population genetics of PAH deficiency, and (2) the study of genotypes, and their various combinations, as a means of explaining and predicting the phenotypic variability observed for the disorders of PAH deficiency.

Introduction Classic phenylketonuria (PKU) is an autosomal recessive human genetic disorder caused by a deficiency of the hepatic enzyme phenylalanine hydroxylase (phenylalanine 4-monooxygenase, E.C. 1.14.16.I), which catalyzes the hydroxylation of phenylalanine to tyrosine by molecular oxygen in the presence of the cofactor tetrahydrobiopterin (Kaufman 1976). This inborn error of amino acid metabolism causes postnatal brain damage and severe mental retardation in untreated affected children. A deficiency in this enzyme results in the excretion of large quantities of phenylpyruvate in the urine, an accumulation of phenylalanine in the blood resulting in hyperphenylalaninemia, and the abnormal formation of a myelin sheath around neuronal axons in the central Offprint requests to: D . S . Konecki

nervous system (Kaufman 1976). While classic PKU is the best known of the hyperphenylalaninemic disorders resulting from PAH deficiency, less severe forms of the disease, mild PKU and hyperphenylalaninemia (HPA), also exist. Clinically and biochemically, these disorders are classified according to pretreatment plasma phenylalanine levels, phenylalanine tolerance (Giittler 1984), residual PAH enzyme activity (determined in liver biopsy material or calculated from phenylalanine tolerance data), and response to protein load (Trefz et al. 1985). PKU is the most common clinically important inborn error of amino acid metabolism, with an average incidence of about 1 in 10000 Caucasian neonates (Bickel et al. 1981). This frequency varies throughout Europe, ranging from approximately 1:2600 in Turkey (Ozalp et al. 1986) to about 1:30000 in Sweden (Veale 1980). In Oriental populations the incidence of PKU ranges from about 1: 16000 in China (Daiger et al. 1989b) to approximately 1:119000 in Japan (Aoki and Wada 1988). These incidence values reflect heterozygote frequencies ranging from about 1:34 in Turkey (this value takes into account the high level of consanguinity reported by Ozalp et al. 1986) to approximately 1: 173 in Japan. The disease state of PKU was first described by Folling in 1934 and identified as an autosomal recessive genetic disorder by Penrose in 1935. Since that time PKU has been the focus of extensive research efforts. Major contributions by Jervis (1947 and 1953), Kaufman (1957, 1958, and 1976) and other biochemists have resulted in the delineation of the complex reaction resulting in the hydroxylation of phenylalanine to tyrosine. In 1954, Bickel et al. introduced a dietary treatment therapy for PKU, which was the first demonstration that hereditary disorders of amino acid metabolism could be corrected by dietary management. The success of this treatment, which must be implemented soon after birth to be effective, provided the impetus to develop a screening method to identify PKU in newborn patients. The resulting bacterial inhibition assay, reported by Guthrie and Susi

378 in 1963, is specific, inexpensive, convenient for analyzing large numbers of samples, and has served as a newborn screening prototype for other metabolic diseases. The newborn screening programs for phenylalanine hydroxylase deficiency have revealed a large variety of clinical and biochemical phenotypes. Only now are we beginning to understand the molecular basis of this phenomenon. The prerequisite for the molecular genetic analysis of the heterogeneity associated with this disease was the isolation of the complementary DNA coding for the PAH enzyme by Prof. S. Woo's group in Houston (Woo et al. 1983). Through the use of in situ hybridization, with a full-length PAH-cDNA as a probe (Kwok et al. 1985), the position of the PAH gene and the PKU locus in man was mapped to the long arm of chromosome 12, position [email protected] (Lidsky et al. 1985a). Eukaryotic, as well as prokaryotic, expression studies with this cDNA have conclusively demonstrated that a single mRNA species, about 2 500 nucleotides in length, contains the genetic information necessary to code for a functional PAH enzyme (Ledley et al. 1985). Subsequently, the results of studies delineating the organization of the PAH gene were published by DiLella et al. (1986a). Based on the observation that specific mutations are in strong linkage disequilibrium with certain haplotypes in [3-thalassemia (Antonarkis et al. 1982; Kazazian et al. 1984), Southern analyses were initiated, using the radiolabelled PAH-cDNA as a probe, to determine the polymorphic nature of the PAH gene. These studies resulted in the establishment of a restriction fragment length polymorphism (RFLP) haplotype-analysis system (Lidsky et al. 1985b; Woo 1988). This system demonstrated conclusively the segregation of mutant PAH genes in the majority of PKU families and is the basis of mutation analysis at this locus. Consequently, this RFLP haplotype-analysis system has been applied to many different populations. Analysis of different mutant alleles has resulted in the detection of some 31 different DNA alterations.

D N A haplotype analysis and ailelic distributions The PAH-cDNA detects ten RFLPs (Lidsky et al. 1985b) at the chromosomal PAH locus, eight of which have been employed for DNA haplotype determination. Currently, more than 46 haplotypes have been observed in the populations studied (Woo 1988; Chen et al. 1989; Daiger et al. 1989a, b; Trefz et al. 1990) out of the 1152 haplotypes theoretically possible for the eight utilized RFLPs, (one three-allelic and seven two-allelic). In the first population investigated, the Danish population, an association between normal and mutant genes and certain haplotypes was observed. About 90% of the PKU chromosomes were associated with four haplotypes, 1 through 4, while haplotypes 2 and 3 were observed predominantly among mutant PAH genes and were rare among normal PAH genes. These two haplotypes, 2 and 3, represented 58% of the mutant PAH genes in Denmark, and 32% were associated with haplotypes 1 and 4.

This led to the hypothesis that there exists an association between certain alleles and specific mutations at the human PAH locus, as had been observed at the 13-globin locus with regard to 13-thalassemias (Orkin and Kazazian 1984). Several reports concerning the haplotype distribution at the human phenylalanine hydroxylase locus for various populations have been published (Chakraborty et al. 1987; Aulehla-Scholz et al. 1988; Herrmann et al. 1988; Lichter-Konecki et al. 1988b; Rey et al. 1988; Riess et al. 1988; Chen et al. 1989; Daiger et al. 1989a, b; Hertzberg et al. 1989; Lichter-Konecki et al. 1989a; Stuhrmann et al. 1989; Sullivan et al. 1989; Apold et al. 1990; Berthelon et al. 1991; Dianzani et al. 1990a, b; Jaruzelska et al. 1991; Svensson et al. 1991; Trefz et al. 1990; Zygulska et al. 1991; U.Lichter-Konecki, unpublished results). These haplotype distributions show certain patterns throughout the different populations. This distribution of the major mutant and normal haplotypes among various populations (and races), as reported in these references, is presented in Fig. 1 and 2, respectively. Essentially all population-based haplotype analyses at this locus (including Caucasian, Euroasian, Asian, and Polynesian) have detected the presence of haplotypes i and 4 among normal (Fig. 2A, B) and mutant PAH alleles (Fig. 1A, B). Populations not following this trend are the Japanese (no normal haplotype 1 alleles detected), Czechoslovakian and Chinese (no mutant haplotype 1 alleles detected in either population). However, relatively few PAH chromosomes have been analyzed for individuals of these three countries. Haplotypes 5 and 7 were the next most frequent normal alleles, while the most frequent mutant alleles were represented by haplotypes 2, 3, 6, and 7. Therefore, haplotypes 1, 4, and 7 have been suggested to be ancient PAH haplotypes predating the divergence of races (Hertzberg et al. 1989), while the other alleles probably emerged from these haplotypes by way of different mutational or recombinational events. The recent identification of two silent (same-sense) mutations (A---~G, Gln232--*Gln and G---~A, Va1245--~Val; Lichter-Konecki et al. 1990) within the coding region of the PAH gene, and subsequent studies to determine haplotype association, demonstrate conclusively that haplotype 7 is derived from haplotype 4. Our studies have found the Gln 232 silent mutation in PAH exon 6 to be associated with normal and mutant haplotype 3, 4, and 7 alleles, whereas the Va1245 same-sense mutation in PAH exon 7 is specifically associated with haplotype 4 alleles. Thus, only haplotypes 1 and 4 can be considered ancient PAH haplotypes. To further test this hypothesis we performed PAH-DNA haplotype analysis of DNA isolated from a chimpanzee and an orang-utan. The chimpanzee shared most, but not all, polymorphic restriction fragments with man, while exhibiting a different pattern of constant bands overall. These results prevented a direct comparison of chimpanzee alleles with human alleles at the PAH locus. The orang-utan possessed a restriction fragment pattern different from humans at its PAH gene locus, indicating variability in the positions of the polymorphic restriction sites between both these species of apes and man (U. Lichter-Konecki, unpublisl~ett data).

379 Norwayln= 8(;,)

100"

80

9

Sweden(n=136~

]

I)enmark(n-66)

[]

Scotlandln=33)

[]

France(n=74) Germany(n=290)

60

"~

[]

Czechoskwakia(n=22)

[]

Poland(n= 130)

[]

tlungary(n=40)

[]

Switzerlandfn=31~)

9

Italy{n=68)

[]

Greece(n=17)

Irrl

Turkeytn=72)

0"

r 4O

20, 0

1

2

3

4

5

7

Haplotypes

A

100-

9

China (n=68)

[ ] Japan (N=35) 8O-

60. 40I.L

20,

0

i

m

1 B

6

m

2

3

4 Haplotypes

As shown in Fig. 1A, a decrease in the frequency of the mutant haplotype 2 allele from east to west has been reported in E u r o p e , the frequencies for this allele being highest in Czechoslovakia (68%) and H u n g a r y (55%), then decreasing from East (54%) to West G e r m a n y (26%) and into France (17.6%). [The data of R e y et al. (1988) m a y not reflect the actual distribution in France, since a significant n u m b e r of North African patients were included in their study.] Also demonstrated in Fig. 1A is a decreasing frequency of mutant haplotype 2 and 3 alleles toward the south of E u r o p e , with an increasing "dominance" of other haplotypes among the mutant alleles, i.e., the haplotype 6 allele in Italy and Turkey (Woo et al. 1991). R F L P haplotype analysis of the Turkish population showed that haplotypes 1 and 4 account for the majority of mutant alleles (52%), whereas haplotypes 2

Fig.lA, B. Distribution of the major mutant PAH haplotypes among different populations. These data are presented as frequencies (%) of the mutant chromosomes analyzed in a particular population. To the right of the graphic representation of these data are the shading reference blocks for the countries whose populations were analyzed. In parenthesis is the number (n) of mutant chromosomes investigated. The references from which this data were assembled are as follows: A Norway (Apold et al. 1990), Sweden (Swensson et al. 1991), Denmark (Chakraborty et al. 1987), Scotland (Sullivan et al. 1989), France (Rey et al. 1988), Germany (Aulehla-Scholz et al. 1988; Lichter-Konecki et al. 1988b; Lichter-Konecki unpublished results), Czechoslovakia (Daiger et al. 1989a), Poland (Jaruzelska et al. 1991; Zygulska et al. 1991), Hungary (Daiger et al. 1989a), Switzerland (Sullivan et al. 1989), Italy (Dianzani et al. 1990b), Turkey (Lichter-Konecki et al. 1989a); B China (Daiger et al. 1989b; Chen et al. 1989), and Japan (Daiger et al. 1989b; Trefz et al. 1990)

and 3 together represented only about 3% of the P A H mutant alleles in Turkey. A b o u t 40% of the chromosomes carrying mutant P A H genes segregate with P A H haplotype 6 alleles, which were very rare a m o n g the normal Turkish chromosomes (Lichter-Konecki et al. 1989a). This haplotype shows a low frequency a m o n g normal or mutant chromosomes in northern and western European, as well as in Asian populations. Since the mutant haplotype 6 allele is relatively frequent in Italy, and shows an increasing frequency in Turkey, it may in the future be found to be as dominant a mutant allele in Mediterranian countries, as is the haplotype 3 mutant allele in northern and western Europe. D u e to the apparent low frequency in the populations of China and J a p a n (approximately 15- to 20-fold less than in most Caucasian populations, T h a l h a m m e r 1975),

380 9

100 -

Norway(n-85) Swcdcn(n=132)

80

60

[]

I)enmark(n=66)

[]

Scothmd(n=31)

[]

France(n=68)

9

Gcrmany(n=188)

[]

Czechoslovakia(n=21)

[]

Poland(n=130)

[]

Hungary(n=38)

[]

Switzerland(n=36)

9

Italy(n=63)

[]

Greece(n=10)

[]

Turkey(n=70)

o

cGJ u_

40.

o .

.

.

1

2

.

Mill .... dillll

.

3

.

4

A

5

6

7

Haplotypes

100I China (n=35) [ ] Japan (n=33)

8O o~

[ ] Polynesia (n=630)

"2 60"

[ ] Southeast Asia (n=74)

LL

40,

0 I

B

2

ulation. To the right of the graphic representation of these data are the shading reference blocks for the countries whose populations were analyzed. In parenthesis is the number (n) of mutant chromosomes investigated. These data were derived from the references listed for Fig. 1A, B and from results reported by Hertzberg et al. (1989) (B)

d

I J

1

3

4 5 Haplotypes

Fig.2A, B. Distribution of the major normal P A H haplotypes among different populations. These data are presented as frequencies (%) of the mutant chromosomes analyzed in a particular pop-

6

7

P K U was once considered to be a disorder of Caucasians. Recent investigations have altered this picture, re"vealing the incidence of P K U in China (Daiger et al. 1989b) and Japan (Aoki and Wada 1988) to be 1 : 16000 and 1:119000, respectively. Haplotype analyses among Asians have shown haplotype 4 to account for 74-80% of the mutant alleles (Fig. 1B). Thus far, no mutant haplotype 2 or 3 alleles have been detected in Japan and only one of each in China (Daiger et al. 1989b). These data indicate that haplotypes 2 and 3, and the mutations associated with them, were not present at all until relatively recently in northern and central European populations. The fifth most frequent mutant allele in the German population, the haplotype 7 PKU allele, was also observed in Denmark, France, Norway, Poland, and Sweden, as well as China and Japan. These haplotype distributions, when viewed with the results of the muta-

tions analyzed, indicate certain patterns concerning the distribution of mutations and the relation between P A H haplotypes and specific mutations in different populations.

Association of mutations and alleles at the P A H gene locus

The original observation at the 13-globin locus that specific mutations were in strong linkage disequilibrium with certain 13-thalassemia haplotypes (Orkin and Kazazian 1984) provided the foundation for the approach taken to analyze mutations in the P A H gene resulting in hyperphenylalaninemia. Having observed that haplotype 3 represented 38% of the mutant alleles and only 3% of the normal alleles in the Danish population (G~itt-

381 Table 1. Mutations in the P A H gene as published up to December 1990 Genotype

Effect of mutation

Associated with RFLP haplotype (in population)

Reference

(1) Donor splice site (2) Mis-sense

G---~A intron 12 C----~T b e x o n 12

Truncated protein a Arg4~ a

3 (European) 2 (European) 1 (French Canadian)

(3) Mis-sense

T--+C exon 9

Leu31k--~Pro a

(4) Mis-sense

G--~Ab exon 7

Glu28~

(5) Non-sense

C---~Tb exon 3

Argm--~Ter a

4 (Chinese)

DiLella et al. (1986b) DiLella et al. (1987) John et al. (1990) Lichter-Konecki et al. (1988a) Riess et al. (1988) Hofman et al. (1989) Lyonnet et al. (1989) Abadie et al. (1989) Okano et al. (1990b) Wang et al. (1989a, 1990b)

(6) Mis-sense (7) Mis-sense

A---~G exon i G---~Ab exon 7

Metl---~Val ArgZ61----~Glna

2 (French Canadian) 1 (European)

(8) Mis-sense

G----~Ab exon 5

ArglSS--~Gln ~

4 (European)

Okano et al. (1990a, c) Okano et al. (1990a, c) Dworniczak et al. (1989)

(9) Non-sense

C---~T b e x o n

ArgZ43---~Ter a

4 (E. Europeans)

Wang et al. (1990a) Lichter-Konecki et al. (1990) Lichter-Konecki et al. (1990)

Type of mutation

7

a

1 (German) 10 (German) 7 (Greek) 38 (N. African, one French) 4 (Algerians) 1 (Caucasian)

John et al. (1989)

(10) Mis-sense

T---~C exon 2

Leu48---~Ser

4 (European)

(11) Mis-sense (12) Non-sense

A---~G exon 6 G---~T exon 7

Glu22a--~Gly Gly272-+Ter

4 (Turkish) 7 (Swedish) 7 (Norwegian) 7 (German)

(13) Deletion

exon 11

Leu264deleted

5 (Swedish)

Svensson et al. (1991)

(14) Mis-sense

T---~G exon 8

Phe249---~Cys a

? (Caucasian)

Okano et al. 1989, 1990c)

(15) Mis-sense

C---~Tb exon 7

Arg2S2---~Trp a

1 (Caucasian)

(16) Mis-sense (17) Mis-sense (18) Mis-sense

C---~Tb exon 7 A---~G exon 6 G--~C b exon 12

ProZ81--~Leua Tyr2~ a Arg4X3---~Proa

1 (+4) (Caucasian) 4 (Chinese) 4 (Chinese)

Abadie et al. (1989) Okano et al. (1989, 1990c) Okano et al. (1989, 1990c)

(19) Mis-sense

A - + G exon 12

Tyr414---~Cysa

4 (Caucasian)

(20) Deletion

5' end of gene

N A c (Scots)

(21) Deletion

exon 3

NA c

Avigad et al. (1990) (Yemenite Jews)

(22) Deletion (23) Mis-sense (24) Silent

3' end of gene T---~C exon 7 A---~G exon 6

N A c (Japanese) Leu255--~Ser Gln232--~Gln

Trefz et al. 1990) John et al. (1990) Lichter-Konecki et al. (1990)

(25) Silent

G---~A exon 7

Va1245---~Val

(26) Silent

A---~T exon 11

Va139%--~Val

(27) Acceptor splice site (28) Mis-sense

G---~A intron 4

Truncated protein

4 (Chinese) 4 (Chinese)

G---~Ab exon 7

ArgZ43---~Gln a

4 (Chinese)

Wang et al. (1990b)

(29) Non-sense (30) Non-sense

G---~A exon 10 C-+A or G? exon 11 exon 3

Trp326---~Ter a Tyr356---~Ter a

4 (Chinese) 4 (Chinese)

Wang et al. (1990b) Wang et al. (1990b)

Ile94---~deleted

2 (Portugese)

Caillaud et al. (1990)

(31) Deletion

?

4, 3, 7 (European, Euroasian, Japanese) 4 (European, Euroasian, Japanese) 4, 3, 16, 17, 28 (German)

Svensson et al. (1990) Apold et al. (1990) Lichter-Konecki (unpublished)

Wang et al. (1989b) Wang et al. (1989b) Okano et al. (1990c) Sullivan et al. (1985)

Lichter-Konecki et al. (1990) Dworniczak et al. (1990) Huang et al. (1991) Wang et al. (1990b)

Designates the study of the mutation by expression analysis b Indicates that the mutation involves a CpG dinucleotide c The abbrevation N A refers to the fact the mutation has not been assigned to a know haplotype due to the alteration of RFLP sites and/or the resulting D N A fragments by the mutation a

382 ler et al. 1987a) a n d was also exclusively a s s o c i a t e d with t h e classic d i s e a s e s t a t e , t h e W o o g r o u p p r o p o s e d t h a t a single m u t a t i o n was l i k e l y to b e r e s p o n s i b l e for P K U in p a t i e n t s c a r r y i n g this m u t a n t allele. T h e s a m e r e a s o n i n g was also a p p l i e d to t h e m u t a n t h a p l o t y p e 2 allele, which s h o w s b o t h a s i m i l a r d i s t r i b u t i o n p a t t e r n a n d a clinical phenotype association. The Woo group subsequently c h a r a c t e r i z e d a m u t a t i o n in the splice d o n o r site of int r o n 12 (SP12, D i L e l l a et al. 1986b), which is a s s o c i a t e d with t h e D a n s i h m u t a n t h a p l o t y p e 3 allele, a n d d e l i n e a t e d a m i s - s e n s e m u t a t i o n in e x o n 12 (Arg4~ Trp, Di L e l l a et al. 1987) a s s o c i a t e d with t h e D a n i s h h a p l o t y p e 2 P K U allele. B a s e d o n t h e h a p l o t y p e 2 (Arg4~ a n d h a p l o t y p e 3 (SP12) m u t a t i o n s , it is p o s s i b l e to i d e n tify m u t a n t P A H g e n e s in 69% o f E a s t G e r m a n P K U p a tients, 58% o f D a n i s h p a t i e n t s , d e c l i n i n g to 42% for W e s t G e r m a n p a t i e n t s . It also p e r m i t s t h e d e t e c t i o n o f a b o u t 1 . 5 % , r a t h e r t h a n 3 % , o f T u r k i s h P K U alleles, since the single m u t a n t h a p l o t y p e 3 allele d e t e c t e d d o e s n o t b e a r t h e SP12 m u t a t i o n . A t the p r e s e n t t i m e , a t o t a l o f 31 d i f f e r e n t m u t a t i o n s in the h u m a n P A H g e n e h a v e b e e n r e p o r t e d (see T a b l e 1). S i x t e e n o f t h e m u t a t i o n s listed in this t a b l e ( n u m b e r s 1 - 4 , 7 - 9 , 1 4 - 1 9 , a n d 2 8 - 3 0 ) h a v e also b e e n s t u d i e d at the e x p r e s s i o n level ( D i L e l l a et al. 1987; M a r v i t et al. 1987; L i c h t e r - K o n e c k i et al. 1988a; O k a n o et al. 1990a, c; W a n g et al. 1990b). T h e t h r e e l a r g e r d e l e t i o n s in the P A H g e n e ( T a b l e 1, n u m b e r s 2 0 - 2 2 ) have b e e n localized to specific r e g i o n s o f t h e h u m a n P A H g e n e , a l t h o u g h t h e greatly altered RFLP patterns accompanying these mut a t i o n s p r e c l u d e t h e i r a s s i g n m e n t to k n o w n h a p l o t y p e s . W h i l e m o s t o f t h e m u t a t i o n s r e p o r t e d to d a t e a r e relatively i n f r e q u e n t , six P K U m u t a t i o n s a s s o c i a t e d with h a p l o t y p e 1 t h r o u g h 4 ( T a b l e 1; n u m b e r s 1, 2, 7, 8, 10, 19) a r e f r e q u e n t a m o n g C a u c a s i a n s . T h e s e D N A altera t i o n s a r e all single b a s e t r a n s i t i o n s r e s u l t i n g e i t h e r in a m i n o acid s u b s t i t u t i o n s o r a b e r r a n t m R N A splicing. To determine the frequency and distribution of these six m o s t f r e q u e n t , a n d t h e two a d d i t i o n a l less f r e q u e n t ( T a b l e 1, nos. 9 a n d 16) m u t a t i o n s a s s o c i a t e d with h a p l o t y p e s 1 t h r o u g h 4 in d i f f e r e n t C a u c a s i a n p o p u l a t i o n s ( W o o et al. 1991), w e p e r f o r m e d allele-specific o l i g o n u c l e o t i d e ( A S O ) h y b r i d i z a t i o n o n p o l y m e r a s e chain reaction ( P C R ) - a m p l i f i e d D N A (Saiki et al. 1985; r e v i e w e d b y V o s b e r g 1989). A t o t a l o f 1 3 2 P K U families f r o m t h r e e d i f f e r e n t p o p u l a t i o n s (99 G e r m a n , 24 T u r k i s h , a n d 9 I t a l i a n f a m i l i e s ) , e a c h h a r b o r i n g at least o n e h a p l o t y p e 1, 2, 3, o r 4 allele, w e r e i n v e s t i g a t e d using m u t a t i o n - s p e cific o l i g o n u c l e o t i d e s as p r o b e s . T h e results o f this invest i g a t i o n a r e p r e s e n t e d in T a b l e s 2 a n d 3. T h e h a p l o t y p e 1 m u t a t i o n in e x o n 7 ( A r g 261 to G l n ) was d e t e c t e d in 31% a n d 40% o f t h e G e r m a n a n d T u r k i s h m u t a n t h a p l o t y p e 1 alleles, r e s p e c t i v e l y , as a g a i n s t 72% o f the m u t a n t h a p l o t p y e 1 alleles in S w i t z e r l a n d a n d 25% of m u t a n t h a p l o t y p e 1 alleles in E u r o p e ( O k a n o et al. 1990a). T h e e x o n 12 m u t a t i o n ( A r g 4~ to T r p ) was f o u n d to b e in close l i n k a g e with m u t a n t h a p l o t y p e 2 alleles in the t h r e e p o p u l a t i o n s i n v e s t i g a t e d , a l t h o u g h it is v e r y r a r e in the T u r k i s h a n d I t a l i a n p o p u l a t i o n s . I n d e p e n d e n t studies b y s e v e r a l g r o u p s h a v e also o b s e r v e d n e a r l y c o m p l e t e linkage d i s e q u i l i b r i u m b e t w e e n t h e Arg4~ T r p mis-sense

Table 2. Frequency of mutations, and their association with DNA haplotypes, in 123 PKU families of two populations. Each family possesses at least one mutant haplotype 1, 2, 3 or 4 allele. This study represents the analysis of 99 German and 24 Turkish PKU kindreds Mutation

German PKU alleles

Turkish PKU alleles

H1

H2

H3

H4

H1

H2

H3

H4

Arg26L--~Gln Pro28L-oLeu Arg4~ SP12 Arg158---~Gln Tyr414-+Cys Arg243---~Ter Leu48---~Ser

12 7 -

53 -

32 -

. 12 9 3 3

8 3 -

1

-

-

-

4

1

2 6

-

. . -

-

-

-

-

-

Na F b(%)

40 20%

53 26%

32 47 16% 23%

20 28%

Aft(%)

48% 100% 100% 58%

55%

TPd (%)

-

.

.

. . -

.

1 2%

.

1 19 2% 26%

100% 100% 63%

64%

34%

a N is the total number of mutant alleles investigated in each population b F refers to the frequency (in %) of the mutant allele in the population c Af is the percentage of PKU alleles possessing characterized PAH mutations d TP is the total percentage of all PKU alleles investigated, found to possess identified mutations in the PAH gene Table 3. Frequency of mutations, and their association with DNA haplotypes, in nine Italian PKU families. At least one mutant haplotype, 1, 2, 3, or 4 allele was possessed by each kindred Mutation Arg261---~Gln Pro28L---~Leu Arg4~ SP12 Arga58--~Gln Tyr414---~Cys ArgZ43---~Ter Leu48---~Ser Na Fb(%)

ItalianPKU alleles H1

H2

H3

H4

1 1 . -

2

-

-

-

-

1 1 1 2

2

0

5

5%

3%

8%

7

41%

.

.

.

a N is the total number of mutant alleles investigated in each respecitve population b F refers to the frequency (in %) of the mutant allele in the population m u t a t i o n a n d P K U h a p l o t y p e 2 alleles in t h e D a n i s h ( D i L e l l a et al. 1987), G e r m a n ( L i c h t e r - K o n e c k i et al. 1988b); F r e n c h ( R e y et al. 1988), Swiss (Sullivan et al. 1989), Scottish (Sullivan et al. 1989) a n d P o l i s h (Jaruz e l s k a et al. 1991; Z y g u l s k a et al. 1991) p o p u l a t i o n s . F o u r studies h a v e d e t e c t e d a l a c k o f a s s o c i a t i o n specificity b e t w e e n the A r g 4~ to T r p P K U mis-sense m u t a t i o n a n d P A H - D N A h a p l o t y p e 2. I n t h e F r e n c h s t u d y ( R e y et al. 1988), all p a t i e n t s with a m u t a n t h a p l o t y p e 2 allele

383 and the phenotype of classic PKU carried the Arg 4~ to Trp mutation at this allele, while patients with less severe phenotypes did not. Whether these patients not exhibiting the expected association between mutation and haplotype were of European or North African ancestry was not indicated in the publication. Only two of four Italian PKU haplotype 2 alleles were found to harbor the Arg4~ mis-sense mutation (Dianzani et al. 1990b). John et al. (1990) have found this mutation in exon 12 to occur on haplotype 1 alleles in two FrenchCanadian PKU families. Recently, a Polish study (Zygulska et al. 1991) has reported the detection of the Arg4~ Trp mutation on a single haplotype 5 PKU allele. This case may represent a new mutation, since a minimum of one recombinational and one mutational event would be required to transfer the Arg 4~ to Trp mis-sense mutation from a PKU haplotype 2 allele to a haplotype 5 allele. The intron 12 splice site mutation (SP12; DiLella et al. 1986b) has been found to be associated with all mutant haplotype 3 alleles analyzed in Denmark (DiLella et al. 1986b), France (Rey et al. 1988), Switzerland (Sullivan et al. 1989) and Scotland (Sullivan et al. 1989). In each of the following populations a single mutant haplotype 3 allele has been detected which has not been associated with the intron 12 splice site mutation: German (Aulehla-Scholz et al. 1988), Swedish (Svensson et al. 1991), Italian (Dianzani et al. 1990b) and Turkish (Konecki et al. 1991). The only PKU haplotype 3 allele analyzed in the Turkish population was associated with the Leu48---> Ser mutation, which was otherwise associated only with mutant haplotype 4 alleles (Konecki et al. 1991). Whether this mutation will be detected among other haplotype 3 alleles in the Mediterranean area, or whether it appeared on a PKU haplotype 3 background as the result of a recombinational or gene conversion event remains to be determined. This mutation (Leu 4s to Ser) in P A H exon 2 has since been detected on 7% of German, 33% of Turkish, and two of five Italian mutant haplotype 4 alleles. The exon 5 mis-sense mutation (Arg 158 to Gln) was found to be associated with 27% of the German, 21% of the Turkish, and one of five Italian haplotype 4 alleles, compared with 33% of the mutant haplotype 4 alleles in the Swiss population and 40% of mutant haplotype 4 alleles in Europe (Okano et al. 1990a). Among 12 mutant German haplotype 7 alleles, 2 show a loss of the B a m H I site in exon 7 of the P A H gene. This was also observed in four of five mutant haplotype 7 alleles in the Swedish population (Svensson et al. 1990) and 82% of PKU haplotype 7 alleles in Norway, which comprise about 20% of the mutant P A H alleles in the Norwegian population (Apold et al. 1990). In the Swedish and Norwegian populations, the loss of the B a m H I site has been shown to be the result of a single base transition generating a termination codon (Arg2'~---> Ter; Svenson et al. 1990). This non-sense mutation, resulting in an inactive and severely truncated form of the P A H enzyme, is responsible for PKU in these patients. Since only 2 of 12 German, 4 of 5 Swedish, and 14 of 17 Norwegian mutant haplotype 7 alleles show the loss of the B a m H I restriction endonuclease recognition site, at least one more mutation must have arisen on the back-

ground of this allele, which is the third most frequent normal allele in most populations. Thus far, five different mutations have been detected on haplotypic backgrounds differing from those on which the mutations were originally characterized: (1) the exon 12 haplotype 2 mutation (Arg 4~ to Trp) on a haplotype 1 background in French Canadians (John et al. 1990), on a haplotype 1 background in French Canadians (John et al. 1990), on a haplotype 44 background in Chinese (Tsai et al. 1990), and on a haplotype 5 background in a Polish individual (Zygulska et al. 1991); (2) the exon 7 haplotype 38 mutation (Glu 2s~ to Lys) on a haplotype 4 background (Abadie et al. 1989) and a haplotype 1 background (Okano et al. 1990b); (3) the exon 7 Pro 281 to Leu mutation on haplotype 1 and 4 backgrounds (Okano et al. 1989); (4) the exon Leu3n-->Pro mutation on haplotype 1, 10, and 7 backgrounds (Lichter-Konecki et al. 1988a; Riess et al. 1988; Hofman et al. 1989); and (5) the exon 2 Leu 4s to Set mutation on haplotype 3 and 4 backgrounds (Konecki et al. 1991). It seems quite remarkable that the exon 12 mis-sense mutation (Arg 4~ to Trp) has now been detected on four haplotypic backgrounds and the exon 7 Glu 2s~ to Lys mutation on three haplotypic backgrounds. While recombinational or gene conversion events must be considered as a means of transferring such mutations from one allele to another, these mechanisms appear relatively unlikely for these two mutations (Arg4~ and GluZS~ Only in the case of the exon 12 mis-sense mutation being transferred from a haplotype 2 allele to a haplotype 1 allele (or vice versa) would a single recombinational event be sufficient to explain such a transfer. A more likely explanation concerns the nature of the bases involved in these two mutations. In both instances (the Arg4~ and the Glu28~ mis-sense mutations), the substituted base is a member of a CpG dinucleotide, which has been shown to be highly mutable, providing a "hot spot" for mutation (Cooper and Youssoufian 1988; Abadie et al. 1989). Thus far, ten different mutations (Table 1, nos. 2, 4, 5, 7-9, 15, 16, 18, and 28) involving nine CpG dinucleotides have been identified. It remains to be determined exactly how many of the 22 CpG dinucleotides contained in the transcribed region encoding the P A H enzyme have incurred mutational alteration.

Conclusions Association between D N A haplotype and mutation

In summary, the initial hypothesis of association between mutations and D N A haplotypes at the P A H locus remains a general rule, the few exceptions having been mentioned in the preceeding paragraph. The distributions of D N A haplotypes in different populations have been observed to vary considerably. Mutant haplotype 2 and 3 alleles are frequent among European populations north of the Alps and there are specific mutations associated with them. By contrast these alleles, and their associated mutations, are of little significance in European populations south of the Alps. A different haplotype 2

384 mutation (Met 1 to Val) has been observed among French Canadian PKU patients (John et al. 1990). Since haplotypes 1 and 4 are found in essentially all populations investigated, with approximately equal frequencies among normal and PKU chromosomes, these haplotypes are considered to be the most ancient alleles. Also, due to their relatively high frequencies, multiple mutations are assumed to have occurred on haplotype 1 and 4 backgrounds. Indeed, five different mutations have already been identified in Caucasian haplotype 4 alleles (Arg 15s to Gln, Dworniczak et al. 1989; Okano et al. 1990a; Glu 2s~ to Gly, Abadie et al. 1989; Pro 281 to Leu, Okano et al. 1989; Arg 243 to Ter, Wang et al. 1990a; Leu 4s to Ser, Konecki et al. 1991) and four different mutations in haplotype 1 alleles (Leu 311 to Pro, Lichter-Konecki et al. 1988a; Olu 28~to Lys, Okano et al. 1990b; Arg 261 to Gln, Okano et al. 1990a; Pro 281 to Leu, Okano et al. 1989; Arg 252 to Trp, Okano et al. 1989). These nine mutations have been observed with different frequencies in different Caucasian ethnic groups, suggesting population-specific patterns. Another nine mutations (Table 1, nos. 5, 17, 18, 22, and 26-30), most of which are tightly linked to P A H - D N A haplotype 4 alleles, are observed only in Oriental populations, supporting the hypothesis of the recent occurrence of PKU mutations on haplotypic backgrounds after divergence of the races (Levy 1989). The only D N A alterations within the region coding for the P A H enzyme which have been detected in Orientals as well as Caucasians are the silent mutations (Gln232--->Gln and ValZ4L--~Val, Lichter-Konecki et al. 1990) residing in exon 6 and 7, respectively. Our laboratory has found both of these same-sense mutations to be associated with mutant as well as normal haplotype 4 alleles, whereas the Gln 232 to Gln silent mutation was also associated with normal and mutant haplotype 3 and 7 alleles in all populations investigated (German, Italian, Turkish, Kuwaiti, and Japanese). Therefore, with few exceptions, distinct mutations are associated with specific D N A haplotypes, although more than one mutation may be associated with a particular halotype. This is especially true with regard to the most frequent haplotypes in different populations. Mutation, and alleles, of the P A H gene also show population-specific distributions.

Origins of the P A H mutations Obviously, there must have been different origins, throughout the world, for the various mutations in the P A H genes which result in disorders of phenylalanine hydroxylade deficiency. Haplotypes 2 and 3, and the mutations associated with them, seem to have occurred relatively recently in European populations north of the Alps, with the highest frequency of mutant haplotype 3 alleles reported for Denmark and of mutant haplotype 2 alleles for Czechoslovakia. Therefore, one may speculate that these might be the populations in which the respective alleles arose and from which they may have spread. Such a theory is not only supported by the gradual decline in the frequency of the haplotype 2 allele from the east to the west of Europe, but also with the ob-

servation of an increased incidence of PKU in northwestern Germany after World War II, resulting from the migration of northeastern German families into this area (Flatz et al. 1984). Evaluation of a substantial collection of European P A H haplotype analysis data (Daiger et al. 1989a) has failed to confirm the earlier suggestion for a Celtic origin of PKU in Europe (Carter and Woolf 1961). The different haplotype 4 mutations obviously have different origins as exemplified by the varying frequencies with which these mutations are observed in different populations. For example, the exon 2 mutation (Leu 48 to Ser; Konecki et al. 1991) is much more frequent in Turkey than among the other Caucasian populations investigated, while the mutation in exon 5 (Arg 158 to Gln) is equally frequent in most Caucasians populations. A third mutation (Arg 243 to Ter in exon 7, Wang et al. 1990a), associated with Caucasian haplotpye 4 alleles, was originally only observed once each in the populations of Hungary and Czechoslovakia, and was subsequently found to occur with similarly low frequencies in our studies of the German, Turkish, and Italian populations. This may, in future, be found to represent a significant mutation in an ethnic group that remains to be analyzed. The haplotype 1 mutation (Arg 261 to Gln), originally characterized in a Swiss individual, was detected in 72% of the mutant haplotype 1 alleles in Switzerland; however, in other European and in Eurasian populations it has been found in lower frequencies (25-40%) of mutant haplotype 1 alleles (Okano et al. 1989; 1990a; this publication). Founder effect has clearly participated in the spread of the deletion of exon 3 among the Yemenite Jewish population (Avigad et al. 1990) and the mutation involving the initiation codon in exon 1 (Met 1 to Val) among FrenchCanadians (John et al. 1989).

Explanations for the high frequency of mutant PAH alleles High mutation rate, hitchhiking of PKU alleles through selection at a nearby locus (interferon Y), founder effect, genetic drift, and natural selection for PKU heterozygotes have all been considered as possible mechanisms for the observed high incidence of PKU (Scriver 1986; Kidd 1987; Scriver et al. 1989). The discovery of two mutations (SP12 and Arg4~ Trp) accounting for more than 50% of the PKU alleles in Denmark (and 42% of those in West Germany) appears to exclude the theory of a high mutation rate at the P A H locus as a cause for the high frequency of PKU. Random genetic drift is excluded, as a general rule, because it would have to affect several different PKU alleles simultaneously in a similar manner. While "founder effect" certainly seems to play a role in the spread of PKU mutations in isolated populations (e.g., in Yemenite Jews and French Canadians), natural selection as the result of a compensating advantage appears to be a more plausible mechanism for the high allele frequencies and the observed high incidence of PKU in many different modern populations (Scriver 1986; Kidd 1987; Woo 1989). The exact nature of such a compensating heterozygote advantage has been a topic of speculation for many years (Saugstad 1976; Woolf 1976,

385 1986; Vogel 1984; Trefz et al. 1989). A recent hypothesis concerns the increased viability of the fetus, afforded by modest hyperphenylalaninemia in the pregnant heterozygote against exposure to ochratoxin A (Woo 1989). This compound is a known ubiquitous mycotoxin abortifacient.

Phenotypic heterogeneity of the hyperphenylalaninemias Taking into consideration the above discussion concerning haplotypes and mutations at the human P A H locus, there are now several explanations for the phenotypic heterogeneity of P A H deficiency at the molecular level. First, it has been shown that different mutations at the P A H locus result in classic P K U (DiLella et al. 1986b; Avigad et al. 1990) or its milder variants (Lyonnet et al. 1989; Okano et al. 1990a, c; Konecki et al. 1991). Most D N A alterations in the P A H gene that have thus far been shown to be associated with phenylalanine hydroxylase deficiency involve single base substitutions and deletions. Therefore, it is clear that no unique D N A alteration of the P A H gene is responsible for this group of metabolic disorders. Second, the majority of Caucasian P K U patients so far analyzed possess two different P A H haplotypes and can be considered haplotypic heterozygotes. Such patients are also compound heterozygotes from the standpoint of mutations, since different haplotypes were shown to be associated with different mutations. After the characterization of the Arg4~ and SP12 mutations (associated with P K U haplotypes 2 and 3, respectively), it became evident that homozygous or exclusively heterozygous (i.e., Arg4~ compound heteroygotes) patients manifest a classic P K U phenotype (Grittier et al. 1987b; H e r r m a n n et al. 1988; Lichter-Konecki et al. 1988b). An even more complex picture emerges from the detection of multiple mutations associated with the mutant haplotype 1 and 4 alleles. The investigation of 138 of our patients (99 German, 24 Turkish, 9 Italian, and 6 others) possessing at least one haplotype 1-4 allele for six relatively frequent (Table 1; nos. 1, 2, 7, 8, 10, and 19) and two less common (Table 1, nos. 9 and 16) mutations allowed complete genotype analysis with regard to the mutations in the P A H genes of 59 individuals. For 52 of these patients, sufficient clinical data (i.e., pretreatment plasma phenylalanine levels and results of protein loading studies) had been collected to permit a correlation of the genotypes with the clinical phenotypes (Okano et al. 1991). The resuits of this correlation are displayed in Table 4. The intron 12 splice site mutation is almost exclusively associated with the classic phenotype and shows complete concordance with mutant P A H haplotype 3 alleles. Only in those patients possessing a second mutant allele coding for an enzyme with very high residual P A H activity i.e., Tyr 414 to Cys; Okano et al. 1990c does the effect of a single haplotype 3 allele bearing the SP12 mutation not result in a severe P K U phenotype. The exon 12 (Arg4~ mis-sense mutation has been shown to be associated with the classic phenotype either in patients homozygous for the mutation or in those whose

Table 4. Correlation between clinical phenotypes and mutations in 52 patients with complete PAH genotype determination; nd, not determined Mutation in PAH gene (lst allele)

Mutation in PHA gene (2nd allele)

RFLP haplotypes

Arg4%-~Trp

Arg4~ 2/2 Pro28L--~Leu 2/1 Arg26L-*Gln 2/1 Arg15S---~Gln 2/4

0/0 0/0 0/30 0/10

I I I I

7 3 3 4

SP12c

SP12 Arg4~ Arg261---~Gln Leu4S---~Ser

0/0 0/0 0/30 0/n.d.

I I I I

3 8 1 1

Leu48---~Ser

Glu221---~Gly 3 d / 4 ArglSS---~Gln 4/4

ND/ND ND/10

I I

1 1

ArglSS---~Gln ArglSS---~Gln 4/4

10/10

I

4

Arg26L--~Cys

Arg261---~Gln 1/1 Tyr414---~Cys 1/4

30/30 30/50

II II

4 2

ProZSL--~Cys

Tyr414---~Cys 1/4

0/50

II

1

Tyr414---~Cys

Arg4~ Arg4~ SP12

4/2 4/2 4/3

50/0 50/0 50/0

I-II II II

2 3 2

Leu48---~Ser

Leu48---~Ser

4/4

ND

II

2

3/3 3/2 3/1 3/4

Residual Clinical No.of activitya pheno- pa% typeb tients

a Residual PAH enzyme activity determined as described by Okano et al. 1990a, c b I, Clasic PKU; II, mild PKU; III, hyperphenylalaninemia (HPA) ~ SP12 refers to the mutation in the splice donor site of intron 12 d The Leu4S--~Seris otherwise only associated with haplotype 4 alleles

other mutant P A H allele carries the intron 12 splice site mutation. While known to occur in tight linkage with P K U haplotype 2 alleles (DiLella et al. 1987), this mutation has also been detected in two other mutant P A H haplotypes (John et al. 1990; Tsai et al. 1990). In combination with a mutant allele coding for an altered P A H enzyme with high residual activity, the Arg 4~ to Trp mutation has also been observed in association with a mild phenytalanine hydroxylase deficiency phenotype (Lichter-Konecki et al. 1988b, 1989b). Both of these mutations affecting P A H exon 12 (SP12 and Arg4~ result in no detectable residual enzyme activity in the heterologous eukaryotic expression systems previously described (DiLella et al. 1987; Marvit et al. 1987; Okano et al. 1990a, c). The exon 7 (ArgZ61--*Gln) mutation associated with haplotype 1 (Abadie et al. 1989) was originally reported not to result in any decrease of P A H enzyme activity (Okano et al. 1990a). However, subsequent expression analyses have revealed a residual activity, of approximately 30% normal, in the assay system (Okano et al. 1990c). Patients homozygous for this mis-sense mutation exhibit the phenotype of mild PKU. The base substitu-

386 tion in codon 158, located in P A H exon 5, is associated with haplotype 4 alleles. The modified P A H enzyme resulting in the substitution of Gln for Arg yielded a P A H enzyme possessing approximately 10% residual activity in the expression system used (Okano et al. 1990a, c) and was associated with classic P K U in patients homozygous for it. The mis-sense mutation Tyr414---~Cys,identified in P A H exon 12 and associated with haplotype 4 alleles, has been shown to yield the highest residual enzyme activity (50% of normal in the assay system) of all D N A alterations so far investigated at the expression level (Okano et al. 1990c). This mutation is associated with the milder phenotypes even when the second allele harbors a mutation resulting in no detectable residual enzyme activity. The haplotype 4 mutation in exon 2 (LeugS---~Ser4s) is associated with mild P K U in these patients who are homozygous for it (Konecki et al. t991).

result in better patient care by improving the predictability of the clinical course (and outcome) of the disease and may allow for the development of new modes of treatment (e.g., enzyme therapy).

Perspectives on the direct detection of P K U mutations

Abadie V, Lyonnet S, Maurin N, Berthelon M, Caillaud C, Giraud F, Mattei JF, Rey J, Rey F, Munnich A (1989) CpG dinucleotides are mutation hot spots in phenylketonuria. Genomics 5 : 963-939 Antonarkis SE, Orkin SH, Kazazian HH, Goff SC, Boehm CD, Waber PG, Sexton JP, Ostrer H, Fairbanks VF, Chakravarti A (1982) Evidence for multiple origins of the 13E-globingene in Southeast Asia. Proc Natl Acad Sci USA 79 : 6608-611 Aoki K, Wada Y (1988) Outcome of the patients detected by newborn screening in Japan. Acta Paediatr 30 : 429-434 Apold J, Eiken HG, Odland E, Fredriksen A, Bakken A, Lorens JB, Boman H (1990) A termination prevalent in Norwegian haplotype no. 7 PKU genes. Am J Hum Genet 47 : 1002-1007 Aulehla-Scholz C, Vorgerd M, Sautter E, Leupold D, Mahlmann R, Ullrich K, Olek K, Horst J (1988) Phenylketonuria: distribution of DNA diagnostic patterns in German families. Hum Genet 78 : 353-355 Avigad S, Cohen BE, Bauer S, Schwartz G, Frydman M, Woo SLC, Niny Y, Shiloh Y (1990) A single origin of phenylketonuria in Yemenite Jews. Nature 344:168-170 Berthelon M, Caillaud C, Rey F, Labrune P, Melle D, Feingold J, Freazal J, Briard M-L, Farriaux J-P, Guibaud P, Journel H, Maurin N, LeMarrec B, Nivelon J-L, Plauchu H, Saudubray J-M, Tron P, Rey J, Munnich A, Lyonnet S (1991) Spectrum of phenylketonuria mutations in western Europe and North Africa, and their relation to polymorphic DNA haplotypes at the phenylalanine hydroxylase locus. Hum Genet 86 : 355-358 Bickel H, Gerrard J, Hickmans EM (1954) The influence of phenylalanine intake on the chemistry and behavior of a phenylketonuric child. Acta Paediatr 43 : 64-77 Bickel H, Bachmann C, Beckers R (1981) Neonatal mass screening for metabolic disorders. Eur J Pediatr 137 : 133-139 Caillaud C, Lyonnet S, Melle D, Rey F, Berthelon M, Vilarinho L, Vaz Osorio R, Rey J, Munnich A (1990) Molecular heterogeneity of mutant haplotype 2 alleles in phenylketonuria. Am J Hum Genet 47: A152 Carter CO, Woolf LI (1961) The birthplaces of parents and grandparents of a series of patients with phenylketonuria in southwest England. Ann Hum Genet 25 : 57-64 Chakraborty R, Lidsky AS, Daiger SP, Gtittler F, Sullivan S, DiLella AG, Woo SLC (1987) Polymorphic DNA haplotypes at the human phenylalanine hydroxylase locus and their relationship with phenylketonuria. Hum Genet 76 : 40-46 Chen S-H, Hsiao K-J, Lin L-H, Liu T-T, Tang R-B, Su T-S (1989) Study of restriction fragment length polymorphisms at the human phenylalanine hydroxylase locus und evaluation of its potential application in prenatal diagnosis of phenylketonuria in Chinese. Hum Genet 81 : 226-230 Cooper DN, Youssoufian H (1988) The CpG dinucleotide and human genetic disease. Hum Genet 78 : 151-155

With the advent of the polymerase chain reaction (PCR) the number of groups involved in the study of mutations and polymorphisms of the P A H gene has increased. Accordingly, the number of D N A alterations reported has also grown. Only four of the fully characterized mutations listed in Table 1 (nos. 1-4) were identified without the use of in vitro D N A amplification. It can be expected that within the near future the human P A H gene will be sequenced in its entirety, although targeted sequencing will probably be directed at regions containing the polymorphic restriction sites currently used for haplotype analysis. Knowledge of the sequences flanking these sites would allow implementation of P C R amplification and A S O hybridization for haplotype analysis and greatly accelerate this laborious process, which forms the basis for the study of mutations at the P K U locus. Most of the D N A alterations so far reported at this genetic locus represent mutations abolishing P A H enzyme activity and resulting in the most severe form of phenylalanine hydroxylase deficiency. Undoubtedly a phase will be entered in which the D N A changes identified will produce more subtle alterations in the enzyme, necessitating the detailed study of the kinetics of mutated enzymes. Fortunately, systems currently exist and are already being implemented for this purpose, which should provide further insight into the regulation and expression of the human P A H gene and its product. As shown in Table 2, about 64% and 34% of the mutations in the G e r m a n and Turkish populations, respectively, can be detected through the use of A S O probes for eight of the most frequent mutations tested to date. Too few Italian P K U families have been studied to assess the effectiveness of these probes for screening purposes. In most populations it is not yet possible to detect more than 50% of the D N A alterations responsible for disorders due to P A H deficiency. Obviously, many more mutations remain to be characterized for better understanding of the phenotypic heterogeneity of disorders caused by a deficiency in the P A H enzyme, as well as a more thorough knowledge of the regulatory and functional domains of that enzyme. Such knowledge should

Acknowledgements. The authors wish to thank Professor F. Vogel for helpful discussions and for critical reading of the manuscript, as well as Dr. F. K. Trefz and M. Schlotter for continuous support and technical assistance. Professor S. L. C. Woo is also thanked for providing the recombinant phPAH247 and for stimulating discussions. This work was supported by grants from the Deutsche Forschungsgemeinschaft: Tr.17512-1 (to Dr.F.K. Trefz and Professor Dr. A. Sippel), Li.375/1-1/1-2 and Li.375/2-1/2-2 (to U.L.-K.). Support from the Fritz Thyssen Foundation is also acknowledged (grant: 1990/1/51, to Professor Dr. H. J. Bremer).

References

387 Daiger SP, Chakraborty R, Reed L, Fekete G, Schuler D, Berenssi G, Nasz I, Brdicka R, Kamaryt J, Pijackova A, Moore S, Sullivan S, Woo SLC (1989) Polymorphic DNA haplotypes at the phenylalanine hydroxylase (PAH) locus in European families with phenylketonuria (PKU) Am J Hum Genet 45 : 310-318 Daiger SP, Reed L, Huang SS, Zeng YT, Wang T, Lo WHY, Okano Y, Hase Y, Fukuda Y, Oura T, Tada K, Woo SLC (1989b) Polymorphic DNA haplotypes at the phenylalanine hydroxylase (PAH) locus in Asian families with phenylketonuria (PKU) Am J Hum Genet 45:319-324 Dianzani I, CamascheUa C, Saglio G, Ferreri GB, Romeo G, Devoto M, Romano C, Cerone R, Giovannini M, Riva E, Trefz FK, Lichter-Konecki U, Woo SLC (1990a) Haplotype distribution and molecular defects of PKU in Italy. J Inher Metab Dis 13 : 292-294 Dianzani I, Devoto M, Camaschella C, Saglio G, Ferrero GB, Cerone R, Romano C, Romeo G, Giovannini M, Riva E, Angeneydt F, Trefz FK, Okano Y, Woo SLC (1990b) Haplotype distribution and molecular defects at the phenylalanine hydroxylase locus in Italy. Hum Genet 86: 69-72 DiLella AG, Kwok SCM, Ledley FD, Marvit J, Woo SLC (1986a) Molecular structure and polymorphic map of the human phenylalanine hydroxylase gene. Biochemistry 25 : 743-749 DiLella AG, Marvit J, Lidsky AS, Gtittler F, Woo SLC (1986b) Tight linkage between a splicing mutation and a specific DNA haplotype in phenylketonuria. Nature 322:799-803 DiLella AG, Marvit J, Brayton K, Woo SLC (1987) An amino acid substitution involved in phenylketonuria is in linkage disequilibrium with DNA haplotype 2. Nature 327: 333-336 Dworniczak B, Aulehla-Scholz C, Horst J (1989) Phenlyketonuria: detection of a frequent haplotype 4 allele mutation. Hum Genet 84: 95-96 Dworniczak B, Aulehla-Scholz C, Horst J (1990) Phenylalanine hydroxylase gene: silent mutation uncovers evolutionary origin of different alleles. Clin Genet 38 : 270-273 Flatz G, Oelbe M, Herrmann H (1984) Ethnic distribution of phenylketonuria in the North German population. Hum Genet 65 : 396-399 FNling A (1934) lSrberAusscheidung von Phenylbrenztraubens~iure in den Harn als Stoffwechselanomalie in Verbindung mit Imbezilit~it. Z Physiol Chem 227 : 169-176 Gtittler F (1984) Phenylketonuria: 50 years since F/511ing'sdiscovery and still expanding our clinical and biochemical knowledge. Acta Paediatr Scand 73:705-716 Gtittler F, DiLella AG, Ledley FD, Lidsky AS, Kwok SCM, Marvit J, Woo SLC (1987a) Molecular biology of phenylketonuria. Eur J Pediatr 146 [Suppl 1] : A5 : A l l Grittier F, Ledley FD, Lidsky AS, DiLella AG, Sullivan SE, Woo SLC (1987b) Correlation between polymorphic DNA haplotypes at phenylalanine hydroxylase locus and clinical phenotypes of phenylketonuria. J Pediatr 110 : 68-71 Guthrie R, Susie A (1963) A simple phenylalanine method for detecting phenylketonuria in large populations of newborn infants. Pediatrics 32: 338-343 Herrmann FH, Wulff K, Wehnert M, Seidlitz G, Gtittler F (1988) Haplotype analysis of classical and mild phenotype of phenylketonuria in the German Democratic Republic. Clin Genet 34 : 176-180 Hertzberg M, Jahromi K, Ferguson V, Dahl HHM, Mercer J, Mickleson KNP, Trent RJ (1989) Phenylalanine hydroxylase gene haplotypes in Polynesians: evolutinary origins and absence of alleles associated with severe phenylketonuria. Am J Hum Genet 44 : 382-387 Hofman KJ, Antonarakis SE, Missiou-Tsangaraki S, Boehm CD, Valle D (1989) Phenylketonuria in the Greek population: haplotype analysis of the phenylalanine hydroxylase gene and identification of a PKU mutation. Mol Biol Med 6:245250 Huang S-Z, Ren Z-R, Zeng Y-T, Woo SLC (1991) PAH 399 GTA(Val)--*GTT(Val), a new silent mutation found in the Chinese. Hum Genet 86 : 305-306

Jaruzelska J, Henriksen KF, Gtittler F, Riess O, Borski K, Blin N, Slomski R (1991) The codon 408 mutation associated with haplotype 2 is predominant in Polish families with phenylketonuria. Hum Genet 86:247-250 Jervis GA (1947) Studies of phenylpyruvic oligophrenia: the position of the metabolic error. J Biol Chem 169: 651-656 Jervis GA (1953) Phenylpyruvic oligophrenia deficiency of phenylalanine-oxidisingsystem. Proc Soc Exp Biol Med 82:514-515 John SWM, Rozen R, Laframboise R, Laberge C, Scriver CR (1989) Novel PKU mutation on haplotype 2 in French-Canadians. Am J Hum Genet 45 : 905-909 John SWM, Rozen R, Scriver CR, Laframboise R, Laberge C (1990) Recurrent mutation, gene conversion, or recombination at the human phenylalanine hydroxylase locus: evidence in French-Canadians and a catalog of mutations. Am J Hum Genet 46 : 970-974 Kaufman S (1957) The enzymatic conversion of phenylalanine to tyrosine. J Biol Chem 226: 511-524 Kaufman S (1958) Phenylalanine cofactor in phenylketonuria. Science 128 : 1506-1508 Kaufman S (1976) Phenylketonuria: biochemical mechanisms. Adv Neurochem 2 : 1-132 Kazazian HH, Orkin SH, Markham AF, Chapman CR, Youssoufian H, Waber PG (1984) Quantification of the close association between DNA haplotypes and specific ~-thalassaemia mutations in Mediterraneans. Nature 310 : 152-154 Kidd KK (1987) Population genetics of a disease. Nature 327: 282283 Konecki DS, Schlotter M, Trefz FK, Lichter-Konecki U (1991) The identification of two mis-sense mutations at the PAH gene locus in a Turkish patient with phenylketonuria. Hum Genet 87 : 389-393 Kwok SCM, Ledley FD, DiLella AG, Robson KJH, Woo SLC (1985) Nucleotide sequence of a full-length complementary DNA clone and amino acid sequence of human phenylalanine hydroxylase. Biochemistry 24 : 556-561 Ledley FD, Grenett HE, DiLella AG, Kwok SCM, Woo SLC (1985) Gene transfer and expression of human phenylalanine hydroxylase. Science 228 : 77-79 Levy HL (1989) Invited editorial: Molecular genetics of phenylketonuria and its implications. Am J Hum Genet 45 : 667-670 Lichter-Konecki U, Konecki DS, DiLella AG, Brayton K, Marvit J, Hahn TM, Trefz FK, Woo SLC (1988a) Phenylalanine hydroxylase deficiency caused by a single base substitution in an exon of the human phenylalanine hydroxylase gene. Biochemistry 27: 2881-2885 Lichter-Konecki U, Schlotter M, Konecki DS, Labeit S, Woo SLC, Trefz FK (1988b) Linkage disequilibrium between mutation and RFLP haplotype at the phenylalanine hydroxylase locus in the German population. Hum Genet 78 : 347-352 Lichter-Konecki U, Schlotter M, Yaylak C, Ozgtic M, Coskun T, Ozalp I, Wendel U, Batzler U, Trefz FK, Konecki D (1989a) DNA haplotype analysis at the phenylalanine hydroxylase locus in the Turkish population. Hum Genet 81 : 373-376 Lichter-Konecki U, Schlotter M, Trefz FK, Konecki D (1989b) Direct detection of a major PKU mutation in the German population after DNA amplification. Eur J Pediatr 149:120123 Lichter-Konecki U, Schlotter M, Trefz FK, Konecki DS (1990) Identification of new mutations at the phenylalanine hydroxylase gene locus. 5th International Congress of Inborn Errors of Metabolsim, Asilomar, abstract W4.4 Lidsky AS, Law ML, Morse HG, Kao F-T, Rabin M, Ruddle FH, Woo SLC (1985a) Regional mapping of the phenylalaninehydroxylase gene and the phenylketonuria locus in the human genome. Proc Nat1 Acad Sci USA 82: 6221-6225 Lidsky AS, Ledley FD, DiLeUa AG, Kwok SCM, Daiger SP, Robson KJH, Woo SLC (1985b) Extensive restriction site polymorphism at the human phenylalanine hydroxylase locus and application in prenatal diagnosis of phenylketonuria. Am J Hum Genet 37 : 619-634

388 Lyonnet S, Caillaud C, Rey F, Berthelon M, Frdzal J, Rey J, Munnich A (1989) Molecular genetics of phenylketonuria in Mediterranean countries: a mutation associated with partial phenylalanine hydroxylase deficiency. Am J Hum Genet 44:511-517 Marvit J, DiLella AG, Brayton K, Ledley FD, Robson KJH, Woo SLC (1987) GT to AT transition at a splice donor site causes skipping of the preceding exon in phenylketonuria. Nucleic Acids Res 15 : 5613-5628 Okano Y, Wang T, Eisensmith RC, Woo SLC (1989) PKU mutations among Caucasians. Am J Hum Genet 45:A211 Okano Y, Wang T, Eisensmith RC, Steinmann B, Gitzelmann R, Woo SLC (1990a) Mis-sense mutations associated with RFLP haplotypes 1 and 4 of the human phenylalanine hydroxylase gene. Am J Hum Genet 46 : 18-25 Okano Y, Wang T, Eisensmith RC, Grittier F, Woo SLC (1990b) Recurrent mutation in the human pbenylalanine hydroxylase gene. Am J Hum Genet 46: 919-924 Okano Y, Gfittler F, Wang T, Eisensmith RC, Dasovich MB, Woo SLC (1990c) Correlation of PKU mutant genotypes and clinical phenotypes in Caucasians. Am J Hum Genet 47 : A164 Okano Y, Eisensmith R, Gtittler E, Lichter-Konecki U, Konecki DS, Trefz FK, Dasovich M, Wang T, Henriksen K, Lou H, Woo SLC (1991) Molecular basis of phenotypic heterogeneity in PKU. N Engl J Med 324 : 1232-1238 Orkin SH, Kazazian HH (1984) The mutation and polymorphism of the human ]3-globin gene and its surrounding DNA. Annu Rev Genet 18:131-171 Ozalp I, Coskun T, Ceyhan M, Tokol S, Oran O, Erdem G, Tekinalp G, Durmus Z, Tarikahya Y (1986) Incidence of phenylketonuria and hyperphenylalaninaemiain a sample of the Turkish newborn population. J Inher Metab Dis 9 [Suppl 2] : 237239 Penrose LS (1935) Inheritance of phenylpyruvic amentia (phenylketonuria) Lancet II : 192-194 Rey F, Berthelon M, Caillaud C, Lyonnet S, Abadie V, BlandinSavoja F, Feingold J, Saudubray J, Fr6zal J, Munnich A, Rey J (1988) Clinical and molecular heterogeneity of phenylalanine hydroxylase deficiencies in France. Am J Hum Genet 43 : 914921 Riess O, Michel A, Speer A, Meiske W, Cobet G, Coutelle C (1988) Linkage disequilibrium between RFLP haplotype 2 and the affected PAH allele in PKU families from the Berlin area of the German Democratic Republic. Hum Genet 78 : 343-346 Saiki RK, Scharf S, Faloona F, Mullis KB, Horn GT, Erlich HA, Arnheim N (1985) Enzymatic amplification of [3-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230:1350-1354 Saugstad LF (1976) Anthropological significance of phenylketonuria and the importance of heterozygote advantage. J Irish Med Assoc 69 : 405-410 Scriver CR (1986) Soundings at the PKU locus. Trends Genet 2: 251-252 Scriver CR, Kaufman S, Woo SLC (1989) The hyperphenylalaninemias. In: Scriver CR, Beaudet AL, Sly WS (eds) The metabolic basis of inherited disease, vol 1. McGraw-Hill, pp 495-546 Stuhrmann M, Reiss O, M6nch E, Kurdoglu G (1989) Haplotype analysis of the phenylalanine hydroxylase gene in Turkish phenylketonuria. Clin Genet 36:117-121 Sullivan SE, Lidsky AS, Brayton K, DiLella AG, King M, Connor F, Woo SLC (1985) Phenylalanine hydroxylase deletion mutant from a patient with classical PKU. Am J Hum Genet 37 : A177 Sullivan SE, Moore SD, Connor JM, King M, Cockburn F, Steinmann B, Gitzelmann R, Daiger SP, Woo SLC (1989) Haplotype distribution of the human phenylalanine hydroxylase locus in Scotland and Switzerland. Am J Hum Genet 44 : 652-659 Svensson E, Andersson B, Hagenfeldt L (1990) Two mutations with- in the coding sequence of the phenylalanine hydroxylase gene. Hum Genet 85 : 300-304 Svensson E, D6beln U von, Hagenfeldt L (1991) Polymorphic DNA haplotypes at the phenylalanine hydroxylase locus and

their relation to phenotype in Swedish phenylketonuria families. Hum Genet 87:11-17 Thalhammer O (coordinator) (1975) Frequency of inborn errors of metabolism, especially PKU, in some representative newborn screening centers around the world. A collaborative study. Hum Genet 30 : 273-286 Trefz FK, Schmidt H, Bartholom6 K, Mahle M, Matthis P, Pecht G (1985) Differential diagnosis and significance of various hyperphenylalaninemias. In: Bickel H, Wachtel U (eds) Inherited diseases of amino acid metabolism. Thieme, Stuttgart New York, pp 86-100 Trefz FK, Lichter-Konecki U, Konecki D (1989) Phenylketonuria. Curr Opinion Pediatr 1 : 421-427 Trefz FK, Yoshino M, Nishiyori A, Aengeneyndt F, SchmidtMader B, Lichter-Konecki U, Konecki DS (1990) RFLP-patterns in Japanese PKU families: new polymorphisms for the mutant phenylalanine hydroxylase gene. Hum Genet 85 : 121122 Tsai TF, Hsiao KJ, Su TS (1990) Phenylketonuria mutation in Chinese haplotype 44 identical with haplotype 2 mutation in northern-European Caucasians. Hum Genet 84 : 409-411 Veale AMO (1980) Screening for phenylketonuria. In: Bickel H, Guthrie R, Hammerson G (eds) Neonatal screening for inborn errors of metabolism. Springer, Berlin Heidelberg New York, pp 7-18 Vogel F (1984) Clinical consequences of heterozygosity for autosoreal-recessive diseases. Clin Genet 25 : 381-415 Vosberg H-P (1989) The polymerase chain reaction: an improved method for the analysis of nucleic acids. Hum Genet 83 : 1-15 Wang T, Okano Y, Eisensmith R, Huang S-Z, Zeng Y-T, Lo WHY, Woo SLC (1989a) Molecular genetics of phenylketonuria in Orientals: linkage disequilibrium between a termination mutation and haplotype 4 of the phenylalanine hydroxylase gene. Am J Hum Genet 45 : 675-680 Wang T, Okano Y, Eisensmith R, Zeng YT, Huang SZ, Lo WHY, Woo SLC (1989b) Molecular genetics of PKU in Orientals. Am J Hum Genet 45 : A228 Wang T, Okano Y, Eisensmith R, Fekete G, Schuler D, Berencsi G, Nasz I, Woo SLC (1990a) Molecular Genetics of PKU in eastern Europe: A nonsense mutation associated with haplotype 4 of the phenylalanine hydroxylase gene. Somat Cell Mol Genet 16 : 85-90 Wang T, Okano Y, Eisensmith RC, Lo WHY, Zeng YT, Huang SZ, Woo SLC (1990b) Molecular basis and population genetics of PKU in the Orient. Am J Hum Genet 47: A241 Woo SLC (1988) Collation of RFLP haplotypes at the phenylalanine hydroxylase (PAH) locus. Am J Hum Genet 43 : 781-783 Woo SLC (1989) Molecular basis and population genetics of phenylketonuria. Biochemistry 28 : 1-7 Woo SLC, Lidsky AS, Gtittler F, Chandra T, Robson KJH (1983) Cloned human phenylalanine hydroxylase gene allows prenatal detection of classical phenylketonuria. Nature 306:151-155 Woo SLC, Okano Y, Dasovich M, Wang, T, GtRtler F, LichterKonecki U, Konecki DS, Swenson E, Hagenfeldt L, Rey F, Rey J, Cockburn F, Conner M, Gitzelmann R, Steinmenn B, Apold J, Eiken HG, Giovanini M, Riva E, Longhi R, Romano V, Cerone R, Naughten ER, Mullins C, Olzap I, Fekete G, Schuler D, Berenssi G, Nasz I, Brdicka R, Kamaryt J, Pijackova A, Cabalska B, Boszkowa K, Schwartz E, Kalimin VN, Eisensmith R (1991) Molecular population dynamics of phenylketonuria among Caucasians: multiple founding populations in Europe. Am J Hum Genet (in press) Woolf LI (1976) A study of the cause of the high incidence of phenylketonuria in Ireland and West Scotland. J Irish Med Assoc 60 : 398-40l Woolf LI (1986) The heterozygote advantage in phenylketonuria. Am J Hum Genet 38 : 773-775 Zygulska M, Eigel A, Aulehla-Scholz C, Pietrzyk JJ, Horst J (1991) Molecular analysis of PKU haplotypes in the population of Southern Poland. Hum Genet (in press)